ThresholdPlot
This function visualizes how classification confidence impacts the assignment of cells to specific cell cycle states. It does so by plotting the distribution of predicted cell cycle states across a range of score thresholds, using a stacked barplot with standardized cell cycle state colors.
At each threshold:
-
Each cell is assigned to the state with the highest prediction score, only if that score exceeds the threshold.
-
Cells with no scores above the threshold are labeled as “Unknown”.
-
The proportion of cells in each state is then calculated and plotted.
ThresholdPlot(seurat_obj, ...)
| Argument | Range of Values | Description |
|---|---|---|
| seurat_obj | NA | A Seurat object containing metadata columns with cell cycle prediction scores for the following states: ‘Neural.G0’, ‘G1’, ‘Late.G1’, ‘S’, ‘S.G2’, ‘G2.M’, ‘M.Early.G1’ |
| … | … | Additional arguments passed using ggplot functions for further customization. |
Value
Returns a ggplot object showing the proportion of cells classified into each cell cycle state at different classification score thresholds. Cells with a maximum score below a given threshold are labeled as “Unknown” for that threshold.
Example
# plot classification confidence after running PredictCellCycle() funtion
ThresholdPlot(seurat_obj)